We've spent the last two pieces on this site making the case for conjoint analysis and Gabor-Granger price tests as go-to methods for pricing decisions — and at Miller Advisors, Gabor-Granger is usually our preferred approach for pricing a single, well-defined add-on. But it isn't the only way to test a price, and it's worth understanding the alternative it's most often compared against: monadic price testing.
Monadic testing is one of the oldest and most straightforward approaches to price research. It won't always be the right tool for a SaaS pricing decision, but it has real strengths, and there are specific situations where it's worth considering over Gabor-Granger.
How Monadic Price Testing Works
In a monadic price test, each respondent evaluates a single offer at a single price. The total sample is split into separate groups, or cells, with each group shown a different price for otherwise the same product or add-on. Researchers then compare purchase interest, intent to buy, or actual choice across the cells to estimate how demand shifts as price changes.
Because each respondent only ever sees one price, they're usually unaware that pricing is the subject of the study at all — which helps produce a relatively unbiased read on the complete offer, uncomplicated by exposure to other price points.
A simple example: if you're testing a new add-on across five price points, you'd recruit five separate respondent groups, show each group only its assigned price, and ask a single purchase-intent question.
| Respondent group | Price shown | Question |
|---|---|---|
| Group A | $49 | Would you purchase this add-on at $49/month? |
| Group B | $99 | Would you purchase this add-on at $99/month? |
| Group C | $129 | Would you purchase this add-on at $129/month? |
| Group D | $179 | Would you purchase this add-on at $179/month? |
| Group E | $199 | Would you purchase this add-on at $199/month? |
After fielding, suppose the five groups produced the following purchase-intent (or actual take-up) results. Multiplying price by take-up rate for each cell gives an estimate of expected revenue per prospect at each price point — a common starting point for identifying an approximate revenue-optimizing price.
So, using the hypothetical results above:
| Price | Take-up | Expected revenue |
|---|---|---|
| $49 | 72% | $35.28 |
| $99 | 58% | $57.42 |
| $129 | 45% | $58.05 |
| $179 | 30% | $53.70 |
| $199 | 23% | $45.77 |
Hypothetical results for illustration only.
In this hypothetical example, $129/month generates the highest expected revenue per prospect — even though it doesn't have the highest take-up rate or the highest price on its own. It's the same trade-off point we raised in our Gabor-Granger piece: the price that maximizes modeled revenue isn't automatically the price that best serves broader adoption or retention goals, and it's worth deciding which objective matters most before the numbers make the decision for you.
Why It Works Well
Monadic testing presents the offer in a relatively natural, undiluted context. Because each respondent evaluates only one price, there's no risk of the anchoring or order effects that can creep into a sequential price ladder — the reaction you get reflects the complete offer, not a comparison against a price the respondent saw a moment earlier.
It's also simple to analyze. Once the cells are fielded, comparing purchase intent or take-up across groups is straightforward, with no need for the more involved trade-off modeling that conjoint analysis requires.
And it's well-suited to validating a small number of already-shortlisted prices for a fixed, well-understood offer — for example, confirming customer reaction to two or three finalist prices that came out of an earlier round of research.
Where It Falls Short
Monadic studies generally require a much larger total sample than Gabor-Granger, because a separate, adequately powered group is needed for every price point tested. Testing five prices means fielding five full cells rather than one ladder — which can make the research considerably more expensive and slower to complete.
The validity of the results also depends on the cells being comparable. If the groups differ in composition — company size, role, usage intensity — those differences can be mistaken for a price effect unless assignment and quotas are carefully managed.
Monadic testing is also less dynamic than either conjoint or Gabor-Granger. It evaluates a fixed, predetermined set of scenarios, and because each respondent only ever sees one price, there's no individual-level price-response curve the way a Gabor-Granger ladder produces — only a comparison of averages across groups.
When to Use Monadic Price Testing
- The product or feature is fixed and already well understood. There's little education burden, so a single, undiluted purchase-intent question is enough.
- You need to validate a small number of already-shortlisted prices, rather than explore a broad range from scratch.
- You can field a large enough sample to power every price cell independently. Budget and reach allow for several full-size groups rather than one shared sample.
- Price isn't normally discussed directly in the purchase process, and you want the cleanest possible read on the complete, undiluted offer.
- You want a second, independent read on a price recommendation that came out of a conjoint or Gabor-Granger study, using a different method as a validation check.
Key Considerations When Designing a Study
- Cell sizes need to be independently powered. Stable estimates typically call for on the order of 100–200+ completes per cell, and that requirement multiplies quickly across price points and segments.
- Random assignment and matched quotas across cells. Keep demographic and firmographic composition comparable cell to cell, so any difference you see reflects price — not who happened to land in which group.
- Screener questions and quota discipline, same as any other pricing study — qualify respondents against your actual buyer profile before they're assigned to a cell.
- Educate before you price. If the product is new or unfamiliar, an uneducated purchase-intent answer isn't meaningful — and unlike a Gabor-Granger ladder, each respondent only gets one shot at the question, so there's less room to recover from early confusion.
- Decide your objective before you see the data. The same revenue-versus-uptake trade-off applies here as with Gabor-Granger — know whether you're optimizing for modeled revenue, adoption, or retention before the results are in front of you.
Monadic Price Testing vs. Gabor-Granger
Both methods test reaction to price. The real difference is in how many prices each respondent sees, and what that trade-off costs you in sample size, bias risk, and the shape of the output.
| Dimension | Monadic Price Testing | Gabor-Granger |
|---|---|---|
| What each respondent sees | One price only | A sequence of prices (a short ladder) |
| Sample size needed | Larger — a separate, independently powered group per price point | Smaller — each respondent contributes data at multiple price points |
| Risk of anchoring / order bias | Low — no exposure to other prices | Present — sequential exposure can anchor responses |
| Output | Purchase intent / take-up at each tested price, compared across groups | A full purchase-intent curve across the tested range from a single sample |
| Best suited to | Validating reaction to a small, already-shortlisted set of prices | Efficiently building a demand curve across a wider price range |
| Field time & cost | Higher, given the larger total sample | Lower, given the smaller sample requirement |
| Miller Advisors' default recommendation | Occasional validation study, or the specific use cases above | Preferred approach for most SaaS add-on and feature pricing decisions |
Gabor-Granger remains our default recommendation for most SaaS add-on and new-feature pricing decisions — it's faster to field, requires a smaller sample, and produces a full demand curve from a single group of respondents. But monadic testing still earns its place on the shelf: when the product is already well understood, when you only need to validate a handful of shortlisted prices, or when you want a second, independent read using a different method, it remains a clean, defensible way to test a price.
